White-tailed Deer ( Odocoileus virginianus ) Subsidize Gray Wolves (Canis lupus ) During a Moose ( Alces americanus ) Decline: A Case of Apparent Competition?

ShAnnon M. B ArBer -M eyer 1, 2, 4 and L. D AviD MeCh 1, 3

1United States Geological Survey, northern Prairie Wildlife research Center, 8711-37 th Street, Se, Jamestown, north Dakota 58401-7317 USA 2United States Geological Survey, 1393 highway 169, ely, Minnesota 55731 USA 3United States Geological Survey, The raptor Center, University of Minnesota, 1920 Fitch Avenue, St. Paul, Minnesota 55731 USA 4Corresponding author: sbarber–[email protected]

Barber-Meyer, Shannon M., and L. David Mech. 2016. White-tailed Deer ( Odocoileus virginianus ) subsidize Gray Wolves (Canis lupus ) during a Moose ( Alces americanus ) decline: a case of apparent competition? Canadian Field–naturalist 130(4): 308–314. Moose ( Alces americanus ) in northeastern Minnesota have declined by 55% since 2006. Although the cause is unresolved, some studies have suggested that Gray Wolves ( Canis lupus ) contributed to the decline. After the Moose decline, wolves could either decline or switch prey. To determine which occurred in our study area, we compared winter wolf counts and summer diet before and after the Moose decline. While wolf numbers in our study area nearly doubled from 23 in winter 2002 to an average of 41 during winters 2011–2013, calf:cow ratios (the number of calves per cow observed during winter surveys) in the wider Moose range more than halved from 0.93 in 2002 to an average of 0.31 during 2011–2013. Compared to summer 2002, wolves in summers 2011–2013 consumed fewer Moose and more White-tailed Deer ( Odocoileus virginianus ). While deer densities were similar during each period, average vulnerability, as reflected by winter severity, was greater during 2011–2013 than 2002, probably explaining the wolf increase. During the wolf increase Moose calves remained a summer food item. These findings suggest that in part of the Moose range, deer subsidized wolf numbers while wolves also preyed on Moose calves. This contributed to a Moose decline and is a possible case of apparent competition and inverse-density-dependent predation. Key Words: Alces americanus ; apparent competition; Canis lupus; diet; Gray Wolf; inverse-density-dependent predation; Min nesota; Moose; Odocoileus virginianus ; scat; White-tailed Deer Introduction source of calf mortality in northeastern Minnesota Gray Wolf ( Canis lupus ) diet in Minnesota generally (Severud et al . 2015). Following the decline in Moose, consists of White-tailed Deer (Odocoileus virginianus ), wolves could either decline or broaden their diet (in - Moose ( Alces americanus) , and Beavers ( Castor can a - crease consumption of an alternate prey) to include deer. densis ; Frenzel 1974; van Ballenberghe et al . 1975; Wolves subsidized by deer could continue to kill Moose Fritts and Mech 1981; Kunkel 1992; Paul 2002). in (even at low Moose densities) potentially resulting in multiple-prey systems, wolf diet may be influenced by an inverse-density-dependent predation rate and appar - “changes in species abundance, prey switching, vegeta - ent competition (holt 1977; holt et al . 1994; Wittmer tion supply, and climatic conditions” (Forbes and The - et al . 2005; hebblewhite and Smith 2010), furthering berge 1996:1512). Wolves respond to declines in pri - the Moose decline. An inverse-density-dependent pre - mary prey by switching to secondary, “buffer” prey dation rate occurs when the predation rate on prey in - (Pimlott et al . 1969; van Ballenberghe et al . 1975; creases while the density of the prey decreases because Mes sier and Crête 1985; Forbes and Theberge 1996). the predator is subsidized by an alternate prey species Moose in northeastern Minnesota declined 55% (to (Wittmer et al . 2005; hebblewhite and Smith 2010). varying degrees in various areas) from a point estimate Apparent competition occurs when two prey species of 8840 in 2006 to 4020 in 2016 (DelGiudice 2016). in directly, negatively interact through the sharing of a Wolves were implicated in this decline based on an in - common predator. in such cases the increasing abun - verse relation between their numbers in the northeastern dance of one prey species indirectly results in the de - part of Moose range and the calf:cow ratio (the num - creasing abundance of the other prey through the nu - be r of calves per cow observed during winter surveys; mer ical response of the shared predator (holt 1977; Mech and Fieberg 2014). Wolves may contribute to holt et al . 1994; Chaneton and Bonsall 2000). Apparent limiting Moose populations (Peterson et al . 1984; Lar - competition has been hypothesized and demonstrated sen et al . 1989) by predation on calves (Testa et al . in various, sympatric ungulate populations including 2000; Bertram and vivion 2002) and were a major Moose, elk (Cervus canadensis ), and Woodland Caribou

308 This work is made available under the Creative Commons CC0 1.0 Universal Public Domain Dedication (CC0 1.0). 2016 BArBer -M eyer AnD MeCh : W oLveS AnD APPArenT CoMPeTiTion 309

(Rangifer tarandus caribou ) living in areas with wolves study area, are more abundant to the northeast in the (Seip 1991, 1992; hurd 1999; Wittmer et al. 2005). larger wolf study area (Frenzel 1974; Mech 2009; Mech Because interactions in large-mammal terrestrial sys - and Fieberg 2014) where Moose may reach densities of tems are particularly complex and because the data re - greater than or equal to 0.2 moose/km 2 (e.g., ≥ 8 moose/ quired to conclude apparent competition are difficult to 34.7 km 2 survey plot) in some locations (DelGiudice acquire, apparent competition is often difficult to distin - 2016). Pre-fawning deer densities (i.e., densities of deer guish in natural food webs from indirect amensalism before fawns are born each year) during 2011–2013 (i.e., when two prey share a common predator, and prey were approximately 1.5–2.0 deer/km 2 (Grund 2014). species “A” is negatively indirectly affected by prey spe - During 2011–2013, we collected wolf scats greater cies “B”, but prey species “B” is not indirectly af fected than or equal to 25 mm wide (to minimize collecting by prey species “A”; Chaneton and Bonsall 2000). scats from smaller sympatric canids such as Coyotes To determine whether wolf abundance declined fol - [Canis latrans ; Weaver and Fritts 1979]) along logging lowing the Moose decline or whether wolves increased roads and trails while conducting other field work that consumption of an alternate prey (i.e., prey switched) in occurred primarily during June–August, but sometimes our study area of northeastern Minnesota (the south - during the fall and winter (Figure 1), in generally the western part of the Mech and Fieberg [2014] study same area as Paul (2002). We collected 38 scats in 2011 , area), we compared wolf numbers and diet in that area 27 in 2012, and 57 in 2013. A portion of some scats before and after the Moose decline. were used to bait traps, and scats were frozen in prepa - ration for analysis. Scats were placed in individual Methods nylon stockings, boiled, rinsed, and then spread out on our study area (“scat study area”; Figure 1) was ap - a plate and allowed to air dry (ibrahim 2015). proximately the southwestern third of a 2060 km 2, long- We analyzed scat contents macro- and microscopi - term wolf study area (Mech 2009) in the east-central cally for deer, Moose, Beaver, Snowshoe hare ( Lepus Superior national Forest (47.8806 on, 91.4162 oW, ap - americanus ), and small mammal remains. The entire proximate centre of our long-term study area) of north - scat contents were examined macroscopically and bones, eastern Minnesota, USA (see nelson and Mech [1981] teeth, claws, feathers, trash, vegetation, soil/rocks, etc. for a detailed description). were recorded. We randomly selected 25 hairs from White-tailed Dee r are more abundant in the scat study each scat using a grid overlain on the spread-out, dried area, whereas Moose, although they inhabit our scat scat (Ciucci et al . 2004; ibrahim 2015) and examined

FiGUre 1. Study area for Gray Wolf ( Canis lupus ) scat collection (Approximate Scat Study Area) during 2011–2013, the long-term wolf study area, the Boundary Waters Canoe Area Wilderness (BWCAW), and Moose ( Alces americanus ) range in northeastern Minnesota, USA. inset shows the state of Minnesota (including Minnesota’s Lake Superior territory) with the main map area of interest shaded.

310 The CAnADiAn FieLD -n ATUrALiST vol. 130

them with a dissecting scope (Swift® Tri–power, San progressed and adult coats were emerging. Thus, great - Jose, CA) at 1–4 magnification. At least one hair rep - er uncertainty exists in August than in our June age- resentative of each species (and age class in ungu - determinations. nevertheless, we are reasonably con - lates) of the 25 hairs was examined (and occasionally fident in our age-class designations because we used negative impressions of the hairs; Kennedy and Carbyn many features to identify them (i.e., characteristics list - 1981) with a compound microscope (Swift® M3500D, ed in the previous sentence). When the weight of evi - San Jose, CA) at four or 10 magnification. We used hair dence was equal for more than one age class, we coded colour, texture, shape, length, diameter, medulla pattern, age class as “unknown”. We did not try to differentiate and scale pattern to identify the prey species, and age juveniles after August. class in ungulates (i.e., fawn or calf versus adult; Ador - We calculated frequency of occurrence of prey cate - jan and Kolenosky 1969; Kennedy and Carbyn 1981; gories and calculated biomass consumed (kg) using Carrlee and horelick 2011; ibrahim 2015). it was gen - prey weights (Table 1) as given by Kunkel (1992) and erally harder to determine ungulate age class as summer Weaver’s (1993) modification of Floyd et al .’s (1978)

TABLe 1. Gray Wolf ( Canis lupus ) scat contents and estimated prey consumption* in the Superior national Forest, Minnesota, USA (2011–2013).

Prey Prey # of scats Frequency # of relative % weight weight containing of Kg % of kg individuals individuals Prey (kg) (kg) / scat the prey item occurrence consumed consumed consumed consumed JUne (PooLeD yeArS ), n scats = 51 Adult deer 58.00 0.9030 14.0 27.5 12.6 40.9 0.2 7.2 Fawn deer 6.30 0.4894 35.0 68.6 17.1 55.4 2.7 90.3 Calf Moose 23.40 0.6262 1.0 2.0 0.6 2.0 0.0 0.9 Beaver 11.25 0.5290 1.0 2.0 0.5 1.7 0.1 1.6 JULy –A UGUST (PooLeD yeArS ), n scats = 55 Adult deer 58.00 0.9030 7.0 10.9 6.3 18.2 0.1 6.3 Fawn deer 13.95 0.5506 30.0 50.0 16.5 47.4 1.2 68.7 Calf Moose 53.10 0.8638 10.5 18.8 9.1 26.1 0.2 9.9 Beaver 11.25 0.5290 5.5 10.9 2.9 8.4 0.3 15.0 2011 ( PooLeD SeASonS ), n scats = 38 Adult deer 58.00 0.9030 11.0 25.0 9.9 40.6 0.2 12.2 Fawn deer 11.40 0.5302 22.0 52.3 11.7 47.6 1.0 73.1 Calf Moose 43.20 0.7846 1.0 2.3 0.8 3.2 0.0 1.3 Beaver 11.25 0.5290 4.0 11.4 2.1 8.6 0.2 13.4 2012 ( PooLeD SeASonS ), n scats = 27 Adult deer 58.00 0.9030 6.0 20.7 5.4 32.5 0.1 8.8 Fawn deer 11.40 0.5302 17.5 62.1 9.3 55.6 0.8 77.0 Calf Moose 43.20 0.7846 0.5 3.5 0.4 2.4 0.0 0.9 Beaver 11.25 0.5290 3.0 10.3 1.6 9.5 0.1 13.3 2013 ( PooLeD SeASonS ), n scats = 57 Adult deer 58.00 0.9030 17.0 28.8 15.4 40.1 0.3 14.9 Fawn deer 11.40 0.5302 25.0 44.1 13.3 34.6 1.2 65.6 Calf Moose 43.20 0.7846 10.0 18.6 7.9 20.5 0.2 10.2 Beaver 11.25 0.5290 3.5 6.8 1.9 4.8 0.2 9.3 ALL yeArS , ALL SeASonS †, n scats = 122 Adult deer 58.00 0.9030 34.0 25.8 30.7 38.5 0.5 12.4 Fawn deer 11.40 0.5302 65.0 50.8 34.5 43.2 3.0 71.1 Calf Moose 43.20 0.7846 11.5 9.9 9.0 11.3 0.2 4.9 Beaver 11.25 0.5290 10.5 9.1 5.6 7.0 0.5 11.6

*Prey weights as used by Kunkel (1992), prey weight per scat (kg) calculated from Weaver’s (1993) equation, # of scats that contained the prey item, frequency of occurrence = # of detections of particular prey item/total prey item detections in all scats (e.g., this included 10 scats that contained two prey items so the total items detected [132] was greater than the total number of scats collected [122]), kg consumed = number of scats × prey weight/scat, number of individuals consumed = kg consumed/average prey weight, and relative percent of individuals consumed = number of individuals consumed of one prey type/the sum of all individuals of all prey types consumed as reflected by the scat contents. in five scats with two primary prey, we considered these as ½ scat for each prey for biomass–consumed calculations. Although we did not consider small mammals as primary prey (four were detected during July–August 2001 and one during July–August 2012), we included them in frequency of occurrence calculations as well as a deer of unknown age class (132 total prey item detections). †All seasons included 17 scats we collected during September–March (13 adult deer and four Beavers). 2016 BArBer -M eyer AnD MeCh : W oLveS AnD APPArenT CoMPeTiTion 311

biomass equation. For scats that contained two primary Results (i.e., not small mammal) prey (5/122 scats or 4%), Winter wolf counts in our scat study area showed we assumed they contained equal amounts of biomass an increase from 23 wolves during 2001–2002 to 45 in consumed (Ciucci et al . 1996) because interior soft tis - 2010–2011, at least 42 during 2011–2012, and a min - sue (e.g., muscle and organs) and hairless portions (e.g., imum (poor survey conditions) of 37 in 2012–2013. Beaver tails and feet) would not be represented by the of 38 scats collected in 2011, 27 in 2012, and 57 in proportion of hair. To better understand the influence of 2013, none contained adult Moose or Snowshoe hare: wolf diet on prey populations, we also calculated the deer were the most common prey (adult 34/122 scats, number of individuals eaten (kg consumed/average fawn 67/122), then calf Moose (13/122) and Beaver prey weight), and the relative percent of individuals (12/122; Table 1). only five of 122 (4%) scats includ - consumed (number of individuals consumed of one ed two primary prey (two fawn deer and Beavers, two prey item/the sum of all individuals of all prey items fawn deer and calf Moose, and one Beaver and calf consumed) as reflected by scat contents by prey type. Moose). Five others (4%) included both fawn deer and To compare our results to those from a previous study small mammal. in one scat, small mammal was the in our area (Paul 2002) that used different prey weights, only prey. we re-calculated Paul’s (2002) biomasses using prey Deer fawns were most frequently detected (51%) and weights given in Kunkel (1992) and Weaver’s (1993) represented the most biomass (43%) and individuals biomass equation. Chi-square tests were used to assess consumed (71%; Table 1). Adult deer were the second differences in the summer counts of prey items between most detected, percent of biomass, and relative individ - the studies (i.e., June–August 2002 and June–August uals consumed (Table 1). Although Beavers and calf 2011–2013; Paul 2002) in Statistix v. 10 (2015). Moose occurred about equally in our scat contents, be - As part of a long-term research project (Mech 2009), cause Beavers are smaller, more individuals were con - wolves were captured and radio-collared (institution - sumed than calf Moose (Table 1). al Animal Care and Use Committee 2015). We located More deer and fewer calf Moose and Beavers were wolves approximately weekly via aerial radio-telemetry detected during June–August 2011–2013 than in June– and calculated winter pack counts for the larger study August 2002 ( χ2 = 19.87, df = 3, P = 0.0002; Tables 1 area as the maximum pack size observed during week - and 2). Because we did not detect any adult Moose, ly locations from December–March each year (Mech we could not compare that category (although Paul 2009). We calculated winter wolf populations using [2002] reported seven scats containing adult Moose only the packs residing in the scat study area (Figure 1). during June 2002 and two during July–August 2002).

TABLe 2. recalculation of prey contents in wolf ( Canis lupus ) scats* for June and July–August 2002 (see Table 4 in Paul [2002]) in part of the Superior national Forest, Minnesota, USA.

Prey Prey # of scats # of relative weight weight containing Kg % of kg individuals % individuals Prey (kg) (kg) / scat the prey item consumed consumed consumed consumed JUNE (2002) † Adult deer 58.00 0.90300 9 8.1 16.3 0.1 4.8 Fawn deer 6.30 0.48940 25 12.2 24.6 1.9 66.3 Adult Moose 227 .00 2.25500 7 15.8 31.7 0.1 2.4 Calf Moose 23.40 0.62620 15 9.4 18.9 0.4 13.7 Beaver 11.25 0.52900 8 4.2 8.5 0.4 12.8 Snowshoe hare 1.08 0.44764 0 0.0 0.0 0.0 0.0 JULy –A UGUST (2002) † Adult deer 58.00 0.90300 3 2.7 7.8 0.1 3.4 Fawn deer 13.95 0.55060 18 9.9 28.5 0.7 52.0 Adult Moose 232.00 2.29500 2 4.6 13.2 0.0 1.5 Calf Moose 53.10 0.86380 16 13.8 39.8 0.3 19.1 Beaver 11.25 0.52900 7 3.7 10.7 0.3 24.1 Snowshoe hare 1.08 0.44764 0 0.0 0.0 0.0 0.0

*Prey weights as used by Kunkel (1992), prey weight per scat (kg) calculated from Weaver’s (1993) equation, # of scats containing the prey item, kg consumed = number of scats × prey weight/scat, number of individuals consumed = kg consumed/ average weight of prey, and relative percent of individuals consumed = number of individuals consumed of one prey type/the sum of all individuals of all prey types consumed as reflected by the scat contents. †During June, two deer were not identifiable to age class and during July–August, two deer and one Moose were not identifiable to age class, so these detections were not included in this comparison analysis. 312 The CAnADiAn FieLD -n ATUrALiST vol. 130

Discussion definitively conclude that apparent competition and The wolf population nearly doubled in our scat study inverse-density-dependent predation of Moose calves area from winter 2002 to 2011, while Moose calf:cow occurred. We do not have precise information on chang - ratios in northeastern Minnesota more than halved from ing prey densities in our scat study area and we did not 0.93 in winter 2002 to 0.24 in 2011, 0.36 in 2012, and evaluate predation rates. We measured wolf diet as re - 0.33 in 2013 (DelGiudice 2016). Although wolf sum - flected by scat contents. But we think apparent com - mer diet in 2011–2013 consisted of fewer Moose and petition and inverse-density-dependent predation like - more deer than summer 2002 prior to the Moose de- ly because 1) wolf populations nearly doubled in our cline (Paul 2002), Moose calves remained a summer study area between the two scat studies presumably food item during 2011–2013. We suspect that wolves increasing predation pressure, 2) calf:cow ratios in the did not decline initially following the Moose decline wider Moose range more than halved between the two be cause they were subsidized, at least temporarily, by studies, and 3) Moose calves remained a summer food deer. item for wolves in the later scat study. Because both Deer densities were low in 2002 and 2011–2013 (i.e., Moose and deer can be negatively, indirectly affected ≤ 2 deer/km 2; Grund 2014), but differences in winter through a shared wolf predator, we do not think this is severity indices (WSi; number of days with minimum a case of indirect amensalism (i.e., where only one of temperature less than or equal to −17.8 oC plus number the two prey species is negatively, indirectly affected of days with greater than or equal to 38 cm snow on the through sharing a common predator; Chaneton and ground: < 100 = mild, 100 –180 = moderate, > 180 = Bonsall 2000). severe; Kohn 1975) among those periods suggest dif - Although our results cannot be generalized to the ferences in deer vulnerability to wolf predation (Mech entire northeastern Minnesota Moose range, wolves and et al . 1971; DelGiudice et al . 2006). The winter before deer inhabit most of that range. our findings might help Paul’s (2002) study was mild (WSi approximately 51– explain the role of wolves and deer in the range-wide 79; Mn Dnr 2015). Winters 2010–2011, 2011–2012 , Moose decline and offer the apparent competition /inverse-density-dependent predation hypothesis to and 2012–2013 included just one mild winter (WSi = be tested further. 120 –180+, 51– 79, and 120 –15 9, respectively; Mn Dnr Another factor (not mutually exclusive) in the Moose 2015). Furthermore, the one “moderate” winter (2012– decline is that deer in our study area could have result - 2013) during our scat study was atypical because the ed in increased incidence of brain worm ( Parelapho - 38 cm snow depth threshold for daily point accumula - strongylus tenuis ) in Moose. This would further alter tion in WSi calculations was not exceeded until mid- Moose vulnerability as well as lead to direct Moose February (Tom rusch, personal communication). Des - mortality from the parasite (Karns 1967; Lankester pite the less than “severe” WSi, a local wildlife manager 2010). Because our study focused mainly on summer reported that “winter (2012–2013) exceeded both 1995– wolf diet (when adult Moose are rarely killed by 1996 and 2013–2014 in deer mortality” (Tom rusch, wolves), we cannot evaluate whether P. tenuis was a personal communication; 1995–1996 and 2013–2014 factor influencing increased wolf predation on Moose. both exceeded WSi of 200 and were considered “sev - Additional research regarding the importance of health- ere”). Thus, although deer densities have remained rela - related factors in the Moose decline and in predispos - tively low between 2002 and 2011–2013, deer were, ing Moose to wolf predation is needed and is ongoing on average, likely more vulnerable to wolf predation by the Minnesota Department of natural resources (M. during our scat study compared to 2002. Carstensen, personal communication). Wolves typically prey on more vulnerable prey (Mech Subsequent to our wolf diet study, wolf populations et al. 1971). When a prey species is more vulnerable in our study area declined (L.D.M. and S.M.B-M., un - (e.g., deep snow would typically hinder deer more than published data). Thus, we suspect that the local deer Moose; Mech et al. 1971), wolves that were more nu - population declined following increased wolf predation. merous due to an alternate prey subsidy could then take Ultimately, it appears that the wolf population eventu - advantage of a new vulnerability in either species. As ally tracked the declining Moose population (L.D.M. well, the average vulnerability of a prey population and S.M.B-M., unpublished data), with a lag due to in - changes during the biological year and subsidized creased deer vulnerability. wolf numbers would have a greater negative impact on We collected scats opportunistically along trails and Moose calf and deer fawn recruitment than if there logging roads. our results may have differed if we had were fewer wolves overall. collected scats near kills (Potvin et al . 1988) or home- our findings suggest that, in at least part of the sites (where adults provision pups). Another regional Moose range, wolves prey-switched and consumed study that examined 1000 wolf scats and included a more deer. At the same time, wolves continued to prey more complex suite of prey (i.e., also adult Moose, on Moose calves, contributing to the Moose decline Black Bear [ Ursus americanus ], Snowshoe hare, and and resulting in a possible case of apparent competi - various small mammals) determined that wolf dietary tion and inverse-density-dependent predation (Wittmer diversity could be determined with as few as 15–50 et al . 2005; hebblewhite and Smith 2010). We cannot scats (ibrahim 2015). Because we considered a more 2016 BArBer -M eyer AnD MeCh : W oLveS AnD APPArenT CoMPeTiTion 313

reduced suite of primary prey our sample likely reflects (Minnesota Department of natural resources, Uni - average summer wolf diet breadth in our scat study versity of Minnesota) and two anonymous reviewers area. Although we did not detect Snowshoe hare and for helpful suggestions on earlier drafts. adult Moose in wolf scats during our study, we expect that wolves did consume them but at such low rates that Literature Cited summer wolf predation would not negatively affect Adorjan, A. S., and G. B. Kolenosky. 1969. A manual for their populations in our study area. the identification of hairs of selected ontario mammals. Scat studies cannot differentiate scavenging and pre - research Branch, ontario Department of Lands and For- dation, and their effect on prey populations might dif - ests, Toronto, ontario, Canada. Bertram, M. R., and M. T. Vivion. 2002. Moose mortality in fer among large ungulate prey species (Forbes and The - eastern interior Alaska. Journal of Wildlife Management 66: berge 1992, 1996). Although biomass calculations are 747–756. less biased than frequency of occurrence (Ciucci et al . Carrlee, E., and L. Horelick. 2011. The Alaska Fur iD Proj - 1996), they do not account for prey condition (poorer ect: a virtual resource for material identification. hiC ob- prey would weigh less), caching, nor incomplete car - jects Specialty Group Postprints 18: 149–171. Ac cessed 7 cass consumption (e.g., bones), and food lost to scav - March 2016. https://alaskafurid.wordpress.com/. engers (Peterson and Ciucci 2003). Chaneton, E. J., and M. B. Bonsall. 2000. enemy-mediated Additional data on seasonal prey densities and sea - apparent competition: empirical patterns and the evidence. oikos 88: 380–394. sonal wolf diet would improve future research. Collect - Ciucci, P., L. Boitani, E. R. Pelliccioni, M. Roco, and I. Guy. ing more scats in the Boundary Waters Canoe Area 1996. A comparison of scat-analysis methods to assess the Wilderness (i.e., northeast of our sampling area; Figure diet of the wolf ( Canis lupus ). Wildlife Biology 2: 37–48. 1), would improve our understanding of wolf diet in an Ciucci, P., E. Tosoni, and L. Boitani. 2004. Assessment of the area with fewer (almost no) deer (nelson and Mech point-frame method to quantify wolf ( Canis lupus ) diet by 2006) and more Beavers (S.M.B-M. and L.D.M., un - scat analysis. Wildlife Biology 10: 149–153. published data). DelGiudice, G. D. 2016. Aerial Moose Survey. Minnesota examining location clusters from GPS-collared Department of natural resources, St. Paul, Minnesota, wolves (Demma et al . 2007) should be used to more USA. DelGiudice, G. D., J. Fieberg, M. R. Riggs, M. C. Powell, precisely determine the influence of wolf predation and W. Pan. 2006. A long-term age-specific survival analy - on Moose and the importance of Moose to the wolf sis of female white-tailed deer. Journal of Wildlife Man - diet. To best determine the influence of wolf predation agement 70: 1556–1568. on Moose, radio-tagged adult and calf Moose should Demma, D. J., S. M. Barber-Meyer, and L. D. Mech. 2007. be studied for cause-specific mortality (Severud et al . Use of GPS telemetry to study wolf predation on deer fawns. 2015). We recommend continued multi-faceted research Journal of Wildlife Management 71: 2767–2775. (e.g., use of thermal refugia, habitat use with respect to Floyd, T. J., L. D. Mech, and P. A. Jordan. 1978. relating deer, etc.) of both radio-tagged adult and calf Moose wolf scat content to prey consumed. Journal of Wildlife to best determine the causes for the recent decline in Management 42: 528–532. Forbes, G. J., and J. B. Theberge. 1992. importance of scav - Moose. enging on moose by wolves in Algonquin Park, ontario. Alces 28: 235–241. Acknowledgements Forbes, G. J., and J. B. Theberge. 1996. response by wolves This study was part of a long-term study supported to prey variation in central ontario. Canadian Journal of by the United States Geological Survey (USGS). We Zoology 74: 1511–1520. thank Dr. M. nelson (USGS retired) for aerial teleme - Frenzel, L. D. 1974. occurrence of moose in food of wolves try prior to 2011 and numerous volunteer technicians as revealed by scat analysis: a review of north American for assisting with scat collection, preparation, and studies. naturaliste Canadien 101: 467–479. Fritts, S. H., and L. D. Mech. 1981. Dynamics, movements, analysis especially W. Binder, C. Brook, S. Malik– and feeding ecology of a newly protected wolf population Wahls, A. Morris, T. roerick, and T. vent. We thank K. in northwestern Minnesota. Wildlife Monographs 80: 1–79. Paul, y. ibrahim, and W. newton for helpful discussions Grund, M. 2014. Monitoring population trends of white-tailed regarding hair analysis and/or statistical analysis. We deer in Minnesota. Farmland Wildlife Populations and re - thank L. Schmidt (international Wolf Center and ver - search Group, Minnesota Department of natural resources, milion Community College) for loaning us the micro - St. Paul, Minnesota, USA. scopes and associated equipment and allowing us to Hebblewhite, M., and D. W. Smith. 2010. Wolf community use classroom space to examine the scats. We thank T. ecology: ecosystem effects of recovering wolves in Banff rusch and Dr. M. Carstensen for personal communica - and yellowstone national Parks. Pages 69–120 in The World of Wolves. Edited by M. Musiani, L. Boitani, and P. tion (both Minnesota Department of natural re sources, C. Paquet. University of Calgary Press, Calgary, Alberta, 2 February and 12 December 2016, respectively). Any Canada. mention of trade, product, or firm names is for descrip - Holt, R. D. 1977. Predation, apparent competition, and the tive purposes only and does not imply en dorsement by struc ture of prey communities. Theoretical Population Biol - the United States Government. We thank W. Severud ogy 12: 197–229. 314 The CAnADiAn FieLD -n ATUrALiST vol. 130

Holt, R. D., J. Grover, and D. Tilman. 1994. Simple rules for Nelson, M. E., and L. D. Mech. 1981. Deer social organiza - interspecific dominance in systems with exploitative and tion and wolf predation in northeastern Minnesota. Wildlife apparent competition. American naturalist 144: 741–771. Monographs 77: 3–53. Hurd, T. E. 1999. Factors limiting moose numbers and their Nelson, M. E., and L. D. Mech. 2006. A 3-decade dearth of interactions with elk and wolves in the Central rocky deer ( Odocoileus virginianus ) in a wolf ( Canis lupus )- Mountains, Canada. Ph.D. thesis, University of British dominated ecosystem. American Midland naturalist 155: Columbia, vancouver, British Columbia, Canada. 373–382. Ibrahim, Y. C. 2015. Seasonal diet composition of gray Paul, K. 2002. Determining summer food habits of 7 wolf wolves ( Canis lupus ) in northeastern Minnesota determined packs in northeastern Minnesota via scat analysis. B.Sc. by scat analysis. M.Sc. thesis, University of Minnesota, thesis, University of Minnesota, St. Paul, Minnesota, USA. Du luth, Minnesota, USA. Peterson, R. O., and P. Ciucci. 2003. The wolf as a carnivore. Institutional Animal Care and Use Committee. 2015. Pages 104–130 in Wolves: Behavior, ecology, and Conser - northern Prairie Wildlife research Center, US Geological vation. Edited by L. D. Mech and L. Boitani. University of Survey, Jamestown, north Dakota, USA. Chicago Press, Chicago, illinois, USA. Karns, P. D. 1967. Pneumostrongylus tenuis in deer in Min - Peterson, R. O., R. E. Page, and K. M. Dodge. 1984. Wolves, nesota and implications for moose. Journal of Wildlife moose and the allometry of population cycles. Science 224: Management 31: 299–303. 1350–1352. Kennedy, A., and L. Carbyn. 1981. identification of wolf Pimlott, D. H., J. A. Shannon, and G. B. Kolenosky. 1969 . prey using hair and feather remains with special reference The ecology of the timber wolf in Algonquin Provincial to Western Canadian national Parks. Canadian Wildlife Park. ontario Department Lands and Forests research Service report, edmonton, Alberta, Canada. report on Wildlife 87: 1–92. Kohn, B. 1975. Winter severity measurements. Wisconsin Potvin, F., H. Jolicoeur, and J. Huot. 1988. Wolf diet and Department of natural resources. Final report, Job 210, prey selectivity during two periods for deer in Quebec: de - Wisconsin, USA. cline versus expansion. Canadian Journal of Zoology 66: Kunkel, K. E. 1992. Factors affecting wolf and black bear 1274–1279. predation of white-tailed deer fawns in northeastern Min - Seip, D. R. 1991. Predation and caribou populations. rangifer nesota. M.Sc. thesis, University of Minnesota, Minneapo - Special issue 7: 46–52. lis, Minnesota, USA. Seip, D. R. 1992. Factors limiting woodland caribou popula - Lankester, M. W. 2010. Understanding the impact of menin - tions and their interrelationships with wolves and moose in geal worm, Parelaphostrongylus tenuis , on moose popu - southeastern British Columbia. Canadian Journal of Zool - lations. Alces 46: 53–70. ogy 70: 1494–1503. Larsen, D. G., D. A. Gauthier, and R. L. Markel. 1989. Severud, W. J., G. DelGiudice, T. R. Obermoller, T. A. Causes and rate of moose mortality in the southwest yukon. Enright, R. G. Wright, and J. D. Forester. 2015. Using Journal of Wildlife Management 53: 548–557. GPS collars to determine parturition and cause-specific Mech, L. D. 2009. Long-term research on wolves in the Su- mortality of moose calves. Wildlife Society Bulletin 39: perior national Forest. Pages 15–34 in recovery of Gray 616–625. Wolf in the Great Lakes region of the United States: an Statistix 10. 2015. Statistix 10. Copyright 1985–2013. Ana - endangered Species Success Story. Edited by A. P. Wyde - lytical Software. version 10.0. Build: 3/6/2015. ven, T. r. vanDeelen, and e. J. heske. Springer, new york, Testa, J. W., E. F. Becker, and G. R. Lee. 2000. Temporal pat - new york, USA. terns in the survival of twin and single moose ( Alces alces ) Mech, L. D., and J. Fieberg. 2014. re-evaluating the north - calves in southcentral Alaska. Journal of Mammalogy 81: eastern Minnesota moose decline and the role of wolves. 162–168. Journal of Wildlife Management 78: 1143–1150. Van Ballenberghe, V., A. W. Erickson, and D. Byman. 1975 . Mech, L. D., L. D. Frenzel, Jr., and P. D. Karns. 1971. The ecology of the timber wolf in northeastern Minnesota. effect of snow conditions on the ability of wolves to cap - Wildlife Monographs 43: 3–43. ture deer. Pages 51–59 in ecological Studies of the Tim - Weaver, J. L. 1993. refining the equation for interpreting ber Wolf in northeastern Minnesota. Edited by L. D. Mech prey occurrence in gray wolf scats. Journal of Wildlife and L. D. Frenzel, Jr. USDA Forest Service research Paper Management 57: 534–538. nC–52 . north Central Forest experimental Station, St. Paul, Weaver, J. L., and S. H. Fritts. 1979. Comparison of coyote Minnesota, USA. and wolf scat diameters. Journal of Wildlife Management Messier, F., and M. Crête. 1985. Moose–wolf dynamics and 43: 786–788. the natural regulation of moose populations. oecologia 65: Wittmer, H. U., A. R. E. Sinclair, and B. N. Mclellan. 503–512. 2005. The role of predation in the decline and extirpation MN DNR [Minnesota Department of Natural Resources]. of woodland caribou. oecologia 144: 257–267. 2015. Winter Severity index (WSi) for White-tailed Deer. Accessed 17 February 2016. http://www.dnr.state.mn.us/ received 21 April 2016 hunting/Deer/maps_stats_archive.html. Accepted 27 September 2016